检测SF6分解特征组分的复合掺杂TiO2纳米管气敏传感器研究

For lack of suitable online monitoring sensor and detection method, online monitoring and insulation fault diagnosis for GIS (gas insulated switchgear) using characteristic component of SF6 decomposition is difficult. In order to solve this problem, TiO2 nanotube based gas sensor array is put forward to detect the types and contents of SF6 decomposition in our application. Studying the mechanism of the composite doped TiO2 nanotube gas responding to the SF6 main decomposition is the key point. Calculate and analyse the interaction and adsorption characteristics between TiO2 nanotube and the selected gas molecules. Propose the adjustment method of band gap so as to enhance its self-cleaning and anti-poisoning characteristic. At the same time, optimize the doping materials and composite doping method of TiO2 to change its fine structure and physicochemical properties to solve the difficulty of its sensitivity to the trace component, selectivity and its life. Build the sensors array which have selectivity to different components and combine it with data fusion and pattern recognition to solve cross sensitivity and the concentration quantity of the component. A large number of experiments are performed in the laboratory research platform using the developed sensor and proposed method to provide key technologies for online monitoring and fault diagnosis of SF6 decomposition.

要实现用SF6分解特征组分含量对气体绝缘设备进行在线监测与绝缘故障诊断，其难点在于缺乏适合在线监测的传感器与检测法。本申请针对这一问题,提出以TiO2纳米管传感器阵列检测SF6分解气体组份种类及含量的方法，重点研究复合掺杂TiO2纳米管对SF6主要分解特征组份的气敏响应机理，计算与分析TiO2纳米管与选定组份分子间的相互作用及吸附特性，提出TiO2纳米管的禁带宽度调整方法，以增强传感器自清洁和抗中毒特性。同时优化掺杂材料和复合掺杂方式，改变其细微结构及理化特性，解决TiO2纳米管传感微量分解组份的灵敏度、选择性和使用寿命的难题。构建具有不同组份选择敏感性的传感器阵列，结合数据融合和模式识别算法，解决对各分解组份交叉敏感及组份含量的定量检测，对研制出的传感器和提出检测方法在实验室研究平台上进行大量试验，为实现用SF6气体分解组份在线监测与故障诊断提供关键技术。

本项目以制备对SF6分解特征组份具有特定敏感性的传感器为实验目标，并通过仿真计算分析不同传感器对SF6分解特征组分的气敏响应机理，主要完成的工作包括：搭建改进了纳米管传感器气敏测试平台以及等离子体改性传感器实验平台；利用阳极氧化法制备了本征TiO2纳米管、脉冲沉积法制备不同量Pt掺杂TiO2纳米管和不同量Au掺杂TiO2纳米管，并测试了这些新型传感器对SF6分解特征组分的气敏响应特性以及传感器的重复使用性、抗中毒性等关键参数；通过Materials Studio软件构建了SF6特征分解组分的气体分子模型，本征锐钛矿TiO2的(101)、(001)晶面完美表面和缺陷表面模型，Pt掺杂锐钛矿TiO2的(101)完美晶面模型以及Au掺杂锐钛矿TiO2的(101)完美晶面模型，分析了TiO2纳米管表面模型与SF6分解组分的气体分子之间的吸附过程、能带结构变化、键能变化、电荷转移、相互作用等；在仿真结果的基础上，分析解释了本征TiO2纳米管的气敏响应机理以及Pt掺杂、Au掺杂对气体吸附行为的影响，更为深入地探究了纳米Pt粒子、Au粒子对TiO2纳米管气敏机理的影响。在研究对SF6分解组分具有高敏感性和特定选择性气敏传感器的基础上，从实用角度出发，进一步研究了基于神经网络的传感器阵列的信息融合方法及气体浓度识别算法。